Pytorch-mighty

Latest version: v0.4.0

Safety actively analyzes 681812 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

0.3.0

* Added `MutualInfoIDTxl` estimator.
* Using native pytorch losses in loss wrappers.
* Added documentation, test coverage.

0.2.1

This release focused on providing support for AutoEncoders.

* Added `MonitorAutoencoder` and `TrainerAutoencoder` classes.
* Fixed bugs in Mutual Information estimators.
* Made `MutualInfoNeuralEstimation` the default MI estimator.

0.1.0

Trainers:
* `TrainerGrad` - a regular gradient descent trainer.

Visualization:
* `Monitor` - a visualization of the training progress. It's used in `TrainerGrad`.
* `MaskTrainer` - shows which input image parts a model "looks at" ([paper](https://arxiv.org/abs/1704.03296)).
* plot adversarial examples with `trainer.train(adversarial=True)`.

Mutual Information estimators:
* `GCMI` - [Gaussian-Copula](https://github.com/robince/gcmi)
* `KMeans` - mutual information binning with k-means (fastest)
* `NPEET` - Kraskov-like kNN mutual information estimation from [NPEET](https://github.com/gregversteeg/NPEET) toolbox
* `MINE` - [Mutual Information Nerual Estimation](https://arxiv.org/abs/1801.04062)

Links

Releases

Has known vulnerabilities

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.